Method and apparatus to provide high precision packet traversal time statistics in a heterogeneous network

Abstract

The present invention relates to handling of performance problems arising in computer networks and in particular, to precise definition of packet traversal time. The basic idea of the present invention comprises the principle to perform a calibration phase, in which the difference between the computer clocks on the client and the server computer is calculated, using a statistical method. This allows to calculate request and response time separately.

1. A method for determining at least one of a request packet traversal time and a response packet traversal time in a computer network between a first computer system and a second computer system comprising:
 determining a clock time difference V between a clock time of the first computer system and a clock time of the second computer system using a statistical method, wherein the clock time of the first computer system and the clock time of the second computer system are unsynchronized; and  
 calculating at least one of the request packet traversal time Dq and the response packet traversal time Dp using said clock time difference V.  
 
   
   
     2. The method according to  claim 1  comprising:
 calculating the request packet traversal time Dq by adding the clock time difference V to an arrival time U 2  at the second computer system of a request packet sent by the first computer system in units of a clock of the second computer and subtracting a transmitting time T 1  at the first computer system of the request packet, as Dq=U 2 +V−T 1 ; and  
 calculating the response packet traversal time Dp by subtracting Dq from D, where D is a time difference in first computer clock units between the transmitting time T 1  of the request packet and a receiving time T 4  at the first computer system of a response packet sent by the second computer system minus a request processing time Ds of the second computer system, as Dp=D−Dq=T 4 −T 1 −Ds−Dq.  
 
   
   
     3. The method according to  claim 1  in which the determining the clock time difference V comprises a repetition of the following:
 predicting a point in time when a request will arrive on the second computer system in units of a clock of the first computer system; and  
 comparing said point in time from the predicting with an actual time of arrival of the request on the second computer system in units of a clock of the second computer system.  
 
   
   
     4. The method according to  claim 3  used for managing the computer network by re-directing packets to a different routing path when at least one of the request packet transversal time and the response packet transversal time indicates a path to be of low performance. 
   
   
     5. The method according to  claim 1  in which the first computer system comprises a client computer system and the second computer system comprises a server computer system. 
   
   
     6. The method according to  claim 1  in which the first computer system comprises a server computer system and the second computer system comprises a client computer system. 
   
   
     7. The method according to  claim 1  in which the first computer system comprises a client computer system and the second computer system comprises a router computer system. 
   
   
     8. The method according to  claim 1  in which the first computer system comprises a router computer system and the second computer system comprises another router computer system. 
   
   
     9. A system for determining at least one of a request packet traversal time and a response packet traversal time in a computer network between a first computer system and a second computer system comprising:
 means for determining a clock time difference V between a clock time of the first computer system and a clock time of the second computer system using a statistical method, wherein the clock time of the first computer system and the clock time of the second computer system are unsynchronized; and  
 means for calculating at least one of the request packet traversal time Dq and the response packet traversal time Dp using said clock time difference V.  
 
   
   
     10. The system according to  claim 9  comprising:
 means for calculating the request packet traversal time Dq by adding the clock time difference V to an arrival time U 2  at the second computer system of a request packet sent by the first computer system in units of a clock of the second computer and subtracting a transmitting time T 1  at the first computer system of the request packet, as Dq=U 2 +V−T 1 ; and  
 means for calculating the response packet traversal time Dp by subtracting Dq from D, where D is a time difference in first computer clock units between the transmitting time T 1  of the request packet and a receiving time T 4  at the first computer system of a response packet sent by the second computer system minus a request processing time Ds of the second computer system, as Dp=D−Dq=T 4 −T 1 −Ds−Dq.  
 
   
   
     11. The system according to  claim 9  in which the means for determining the clock time difference V comprises a repetition of the following:
 predicting a point in time when a request will arrive on the second computer system in units of a clock of the first computer system; and  
 comparing said point in time from the predicting with an actual time of arrival of the request on the second computer system in units of a clock of the second computer system.  
 
   
   
     12. The system according to  claim 11  used for managing the computer network by re-directing packets to a different routing path when at least one of the request packet transversal time and the response packet transversal time indicates a path to be of low performance. 
   
   
     13. The system according to  claim 9  in which the first computer system comprises a client computer system and the second computer system comprises a server computer system. 
   
   
     14. The system according to  claim 9  in which the first computer system comprises a server computer system and the second computer system comprises a client computer system. 
   
   
     15. The system according to  claim 9  in which the first computer system comprises a client computer system and the second computer system comprises a router computer system. 
   
   
     16. The system according to  claim 9  in which the first computer system comprises a router computer system and the second computer system comprises another router computer system. 
   
   
     17. A computer program comprising code portions adapted for performing a method for determining at least one of a request packet traversal time and a response packet traversal time in a computer network between a first computer system and a second computer system comprising:
 determining a clock time difference V between a clock time of the first computer system and a clock time of the second computer system using a statistical method, wherein the clock time of the first computer system and the clock time of the second computer system are unsynchronized; and  
 calculating at least one of the request racket traversal time Dq and the response packet traversal time Dp using said clock time difference V.  
 
   
   
     18. At least one program storage device readable by a machine, tangibly embodying at least one program of instructions executable by the machine to perform a method for determining at least one of a request packet traversal time and a response packet traversal time in a computer network between a first computer system and a second computer system comprising:
 determining a clock time difference V between a clock time of the first computer system and a clock time of the second computer system using a statistical method, wherein the clock time of the first computer system and the clock time of the second computer system are unsynchronized; and  
 calculating at least one of the request packet traversal time Dq and the response packet traversal time Dp using said clock time difference V.  
 
   
   
     19. The at least one program storage device of  claim 18 , wherein said method further comprises:
 calculating the request packet traversal time Dq by adding the clock time difference V to an arrival time U 2  at the second computer system of a request packet sent by the first computer system in units of a clock of the second computer and subtracting a transmitting time T 1  at the first computer system of the request packet, as Dq=U 2 +V−T 1 ; and  
 calculating the response packet traversal time Dp by subtracting Dq from D, where D is a time difference in first computer clock units between the transmitting time T 1  of the request packet and a receiving time T 4  at the first computer system of a response packet sent by the second computer system minus a request processing time Ds of the second computer system, as Dp=D−Dq=T 4 −T 1 −Ds−Dq.  
 
   
   
     20. The at least one program storage device of  claim 18  in which the determining the clock time difference V comprises a repetition of the following:
 predicting a point in time when a request will arrive on the second computer system in units of a clock of the first computer system; and  
 comparing said point in time from the predicting with an actual time of arrival of the request on the second computer system in units of a clock of the second computer system.  
 
   
   
     21. The at least one program storage device of  claim 20  used for managing the computer network by re-directing packets to a different routing path when at least one of the request packet transversal time and the response packet transversal time indicates a path to be of low performance. 
   
   
     22. The at least one program storage device of  claim 18  in which the first computer system comprises a client computer system and the second computer system comprises a server computer system. 
   
   
     23. The at least one program storage device of  claim 18  in which the first computer system comprises a server computer system and the second computer system comprises a client computer system. 
   
   
     24. The at least one program storage device of  claim 18  in which the first computer system comprises a client computer system and the second computer system comprises a router computer system. 
   
   
     25. The at least one program storage device of  claim 18  in which the first computer system comprises a router computer system and the second computer system comprises another router computer system.